cerl: Make sure that we preserve the invariants for maps

Maps have certain invariants that must be preserved:

(1) A map as a pattern must be represented as #c_map{} record,
never as a literal. The reason is that the pattern '#{}' will
match any map, not just the empty map. The literal '#{}' will
only match the empty map.

(2) In a map pattern, the key must be a literal, a variable, or
data (list or tuple). Keys that are binaries or maps *must* be
represented as literals.

(3) Maps in expressions should be represented as literals if possible.
Nothing is broken if this invariant is broken, but the generated
code will be less efficient.

To preserve invariant (1), cerl:update_c_map/3 must never collapse
a map to a literal. To preserve invariant (3), cerl:update_c_map/3
must collapse a map to a literal if possible.

To preserve both invariants, we need a way for cerl:update_c_map/3 to
know whether the map is used as a pattern or as an expression. The
simplest way is to have an 'is_pat' boolean in the #c_map{} record
which is set when a #c_map{} record is initially created.

We also need to update core_parse.yrl to establish the invariants
in the same way as v3_core, to ensure that compiling from a
.core file will work even if all optimizations on Core Erlang are
disabled.

4 files changed, 78 insertions, 17 deletions

diff --git a/lib/compiler/src/cerl.erl b/lib/compiler/src/cerl.erl
index 7a2c3d70de..96c8e02542 100644
--- a/lib/compiler/src/cerl.erl
+++ b/lib/compiler/src/cerl.erl
@@ -123,11 +123,13 @@
 	 bitstr_flags/1,
 
 	 %% keep map exports here for now
+	 c_map_pattern/1,
 	 map_es/1,
 	 map_arg/1,
 	 update_c_map/3,
 	 c_map/1, is_c_map_empty/1,
 	 ann_c_map/2, ann_c_map/3,
+	 ann_c_map_pattern/2,
 	 map_pair_op/1,map_pair_key/1,map_pair_val/1,
 	 update_c_map_pair/4,
 	 c_map_pair/2,
@@ -1590,7 +1592,17 @@ map_arg(#c_map{arg=M}) ->
 -spec c_map([c_map_pair()]) -> c_map().
 
 c_map(Pairs) ->
-    #c_map{es=Pairs}.
+    ann_c_map([], Pairs).
+
+-spec c_map_pattern([c_map_pair()]) -> c_map().
+
+c_map_pattern(Pairs) ->
+    #c_map{es=Pairs, is_pat=true}.
+
+-spec ann_c_map_pattern([term()], [c_map_pair()]) -> c_map().
+
+ann_c_map_pattern(As, Pairs) ->
+    #c_map{anno=As, es=Pairs, is_pat=true}.
 
 -spec is_c_map_empty(c_map() | c_literal()) -> boolean().
 
@@ -1598,9 +1610,9 @@ is_c_map_empty(#c_map{ es=[] }) -> true;
 is_c_map_empty(#c_literal{val=M}) when is_map(M),map_size(M) =:= 0 -> true;
 is_c_map_empty(_) -> false.
 
--spec ann_c_map([term()], [cerl()]) -> c_map() | c_literal().
+-spec ann_c_map([term()], [c_map_pair()]) -> c_map() | c_literal().
 
-ann_c_map(As,Es) ->
+ann_c_map(As, Es) ->
     ann_c_map(As, #c_literal{val=#{}}, Es).
 
 -spec ann_c_map([term()], c_map() | c_literal(), [c_map_pair()]) -> c_map() | c_literal().
@@ -1648,10 +1660,12 @@ fold_map_pairs(As,[#c_map_pair{op=#c_literal{val=exact},key=Ck,val=Cv}=E|Es],M)
 fold_map_pairs(As,Es,M) ->
     #c_map{arg=#c_literal{val=M,anno=As}, es=Es, anno=As }.
 
-%-spec update_c_map(c_map() | c_literal(), [c_map_pair()]) -> c_map() | c_literal().
+-spec update_c_map(c_map(), cerl(), [cerl()]) -> c_map() | c_literal().
 
-update_c_map(Old,M,Es) ->
-    #c_map{arg=M, es = Es, anno = get_ann(Old)}.
+update_c_map(#c_map{is_pat=true}=Old, M, Es) ->
+    Old#c_map{arg=M, es=Es};
+update_c_map(#c_map{is_pat=false}=Old, M, Es) ->
+    ann_c_map(get_ann(Old), M, Es).
 
 map_pair_key(#c_map_pair{key=K}) -> K.
 map_pair_val(#c_map_pair{val=V}) -> V.
diff --git a/lib/compiler/src/core_parse.hrl b/lib/compiler/src/core_parse.hrl
index 4a00535360..7fd4a82e4e 100644
--- a/lib/compiler/src/core_parse.hrl
+++ b/lib/compiler/src/core_parse.hrl
@@ -72,7 +72,8 @@
 
 -record(c_map, {anno=[],
 		arg=#c_literal{val=#{}} :: cerl:c_var() | cerl:c_literal(),
-		es :: [cerl:c_map_pair()]}).
+		es :: [cerl:c_map_pair()],
+		is_pat=false :: boolean()}).
 
 -record(c_map_pair, {anno=[],
 	             op :: #c_literal{val::'assoc'} | #c_literal{val::'exact'},
diff --git a/lib/compiler/src/core_parse.yrl b/lib/compiler/src/core_parse.yrl
index a66ad4235f..716c03bc94 100644
--- a/lib/compiler/src/core_parse.yrl
+++ b/lib/compiler/src/core_parse.yrl
@@ -182,14 +182,14 @@ atomic_pattern -> atomic_literal : '$1'.
 tuple_pattern -> '{' '}' : c_tuple([]).
 tuple_pattern -> '{' anno_patterns '}' : c_tuple('$2').
 
-map_pattern -> '~' '{' '}' '~' : #c_map{es=[]}.
+map_pattern -> '~' '{' '}' '~' : c_map_pattern([]).
 map_pattern -> '~' '{' map_pair_patterns '}' '~' :
-		   #c_map{es=lists:sort('$3')}.
+		   c_map_pattern(lists:sort('$3')).
 
 map_pair_patterns -> map_pair_pattern : ['$1'].
 map_pair_patterns -> map_pair_pattern ',' map_pair_patterns : ['$1' | '$3'].
 
-map_pair_pattern -> '~' '<' anno_pattern ',' anno_pattern '>' :
+map_pair_pattern -> '~' '<' anno_expression ',' anno_pattern '>' :
 			#c_map_pair{op=#c_literal{val=exact},key='$3',val='$5'}.
 
 cons_pattern -> '[' anno_pattern tail_pattern :
@@ -284,10 +284,10 @@ tail_literal -> ',' literal tail_literal : #c_cons{hd='$2',tl='$3'}.
 tuple -> '{' '}' : c_tuple([]).
 tuple -> '{' anno_expressions '}' : c_tuple('$2').
 
-map_expr -> '~' '{' '}' '~' : #c_map{es=[]}.
-map_expr -> '~' '{' map_pairs '}' '~' : #c_map{es='$3'}.
-map_expr -> '~' '{' map_pairs '|' variable '}' '~' : #c_map{arg='$5',es='$3'}.
-map_expr -> '~' '{' map_pairs '|' map_expr '}' '~' : #c_map{arg='$5',es='$3'}.
+map_expr -> '~' '{' '}' '~' : c_map([]).
+map_expr -> '~' '{' map_pairs '}' '~' : c_map('$3').
+map_expr -> '~' '{' map_pairs '|' variable '}' '~' : ann_c_map([], '$5', '$3').
+map_expr -> '~' '{' map_pairs '|' map_expr '}' '~' : ann_c_map([], '$5', '$3').
 
 map_pairs -> map_pair : ['$1'].
 map_pairs -> map_pair ',' map_pairs : ['$1' | '$3'].
@@ -307,7 +307,7 @@ tail -> '|' anno_expression ']' : '$2'.
 tail -> ',' anno_expression tail : c_cons('$2', '$3').
 
 binary -> '#' '{' '}' '#' : #c_literal{val = <<>>}.
-binary -> '#' '{' segments '}' '#' : #c_binary{segments='$3'}.
+binary -> '#' '{' segments '}' '#' : make_binary('$3').
 
 segments -> segment ',' segments : ['$1' | '$3'].
 segments -> segment : ['$1'].
@@ -410,9 +410,55 @@ Erlang code.
 
 -include("core_parse.hrl").
 
--import(cerl, [c_cons/2,c_tuple/1]).
+-import(cerl, [ann_c_map/3,c_cons/2,c_map/1,c_map_pattern/1,c_tuple/1]).
 
 tok_val(T) -> element(3, T).
 tok_line(T) -> element(2, T).
 
+%% make_binary([#c_bitstr{}]) -> #c_binary{} | #c_literal{}
+%%  Create either #c_binary{} or #c_literal{} from the binary segments.
+%%  In certain contexts, such as keys for maps, only literals and
+%%  variables are allowed, so we must not create a #c_binary{}
+%%  record in those situation.
+%%
+%%  To keep this function simple, we use a crude heuristic. We will
+%%  assume that Core Erlang has been produced by core_pp. If the
+%%  segments *could* have been output from a literal binary by
+%%  core_pp, we will create a #c_literal{}. Otherwise we will create a
+%%  #c_binary{} record.
+
+make_binary(Segs) ->
+    try make_lit_bin(<<>>, Segs) of
+	Bs when is_bitstring(Bs) ->
+	    #c_literal{val=Bs}
+    catch
+	throw:impossible ->
+	    #c_binary{segments=Segs}
+    end.
+
+make_lit_bin(Acc, [#c_bitstr{val=I0,size=Sz0,unit=U0,type=Type0,flags=F0}|T]) ->
+    I = get_lit_val(I0),
+    Sz = get_lit_val(Sz0),
+    U = get_lit_val(U0),
+    Type = get_lit_val(Type0),
+    F = get_lit_val(F0),
+    if
+	is_integer(I), U =:= 1, Type =:= integer, F =:= [unsigned,big] ->
+	    ok;
+	true ->
+	    throw(impossible)
+    end,
+    if
+	Sz =< 8, T =:= [] ->
+	    <<Acc/binary,I:Sz>>;
+	Sz =:= 8 ->
+	    make_lit_bin(<<Acc/binary,I:8>>, T);
+	true ->
+	    throw(impossible)
+    end;
+make_lit_bin(Acc, []) -> Acc.
+
+get_lit_val(#c_literal{val=Val}) -> Val;
+get_lit_val(_) -> throw(impossible).
+
 %% vim: syntax=erlang
diff --git a/lib/compiler/src/v3_core.erl b/lib/compiler/src/v3_core.erl
index b861de3bbb..1d5eb85a1c 100644
--- a/lib/compiler/src/v3_core.erl
+++ b/lib/compiler/src/v3_core.erl
@@ -1638,7 +1638,7 @@ pattern({tuple,L,Ps}, St) ->
     {annotate_tuple(record_anno(L, St), Ps1, St),Eps,St1};
 pattern({map,L,Pairs}, St0) ->
     {Ps,Eps,St1} = pattern_map_pairs(Pairs, St0),
-    {#c_map{anno=lineno_anno(L, St1), es=Ps},Eps,St1};
+    {#c_map{anno=lineno_anno(L, St1),es=Ps,is_pat=true},Eps,St1};
 pattern({bin,L,Ps}, St) ->
     %% We don't create a #ibinary record here, since there is
     %% no need to hold any used/new annotations in a pattern.